Inaccurate Listing Specs Drive Up Returns After AI RecommendationsBrief · digest for November 7
Brief≈ 1 min

Inaccurate Listing Specs Drive Up Returns After AI Recommendations

Deploying smart conversational assistants in online retail has brought sellers an unexpected problem: a sharp rise in returns. According to research by analytics platforms Narvar and Blue Yonder, 46% of shoppers who ordered items recommended by generative AI assistants returned at least one purchase over the year.

At the same time, the breakdown of return reasons flipped. While wrong size topped the list in traditional e-commerce (around 42% of returns), 36% of AI-driven returns were caused by direct mismatch with the product description.

The mechanics are simple: when a buyer asks AI to find an item with exact parameters (say, a lamp strictly under 35 cm with an E27 socket) and the seller left structured fields blank in the dashboard, the model guesses specs from broad niche patterns. The shopper trusts the assistant's confident answer, places an order, and receives an incompatible item — leaving the seller to pay for return shipping and lose listing rankings.

In our piece on reducing marketplace return rates, we emphasized that the era of vague copy is over. To make AI bring buyers without returns, fill in 100% of formal specification fields.

Want a deep diveVotes help us choose what to break down in detail. This material has 0 votes.
Sources · 2